๐Ÿ— KeyzHub
19Keys ยท community archive
6102 bytes raw
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
# Playbook: Nine-Word Reactivation

**Deliverable type this produces:** `reactivation-campaign`
**Derives from:** Dean Jackson โ€” direct-response marketer credited with the "nine-word email," a single plain-text question sent to dead/cold segments (e.g., "Are you still looking to sell your house?"). Named for its typical length, though the exact word count varies; the doctrine is brevity and a bare question, not a fixed nine.
**Engine (the science):** The mechanism is conversational/curiosity-based rather than persuasion-based โ€” a short, personal-sounding, plain-text question triggers a reply because it reads as a real human asking a real question, not a broadcast. It produces the highest reply-rate-per-word of any email format because it removes every element (design, links, offers) that signals "marketing email" and could be ignored on sight. It also functions as a deliverability-protective mechanism: replies are the strongest positive engagement signal to mailbox providers, and this format is engineered to maximize replies from a segment that would otherwise sit silent or get suppressed.

## When to use it

- Cold segments: 90-180+ days with no open/click, or previously-purchasing customers who've gone quiet (Layer 2's "gone quiet" lifecycle stage).
- Before sunsetting a segment โ€” this is the last genuine attempt to determine if a contact is still alive/interested before suppression.
- For 19Keys: any slice of the 4,591-buyer list that doesn't respond to the initial re-permission warm-up wave (ยง9.6 step 4) โ€” this is the fallback for the ones the daily/welcome cadence can't wake.

## When NOT to use it

- On already-engaged (hot/warm) segments โ€” it "feels robotic" to someone who already interacts regularly (ยง4.10 explicit failure case); save the format's power for genuinely cold contacts.
- With no sunset plan behind it โ€” sending a bare reactivation question and then doing nothing with non-responders wastes the signal it's designed to produce. This playbook only works as the second-to-last step before suppression, not a standalone tactic.
- As a disguised sales pitch โ€” if the "question" secretly links to an offer or reads as a setup for a pitch, it breaks the format's entire mechanism (it stops reading as a real question) and will underperform *and* damage trust.

## The step-by-step build

1. **Define the target segment by an explicit inactivity rule.** Not "people who seem cold" โ€” a stated rule: e.g., "no open or click in 90 days" or "no purchase in 180 days despite prior purchase history." This explicit rule is required by the verify gate.
2. **Write email 1 as a single plain-text question, โ‰ค25 words total** (subject + body combined is the practical constraint the rubric enforces). No links, no images, no CTA button โ€” just a first-name-personalized, real question relevant to why they originally engaged. Example pattern: "Hey [First] โ€” still working on [X]? No worries either way, just closing out my list."
3. **Wait for replies.** This email is designed to be replied to directly, not clicked. Route replies to a real inbox/person, not an autoresponder.
4. **Write email 2 (of a 2-3 email arc)** for non-responders โ€” a slightly warmer, still-brief follow-up, still no hard pitch, still asking rather than telling. Optional email 3 for a final soft check-in if the segment's value justifies a third touch.
5. **Define the sunset rule explicitly.** What happens to non-responders after the arc completes โ€” e.g., "move to suppressed/dead segment; no further sends without a fresh explicit opt-in." Missing this is an automatic fail (ยง8.1) โ€” a reactivation campaign with no sunset rule just becomes indefinite cold-mailing, which is the exact behavior this playbook exists to prevent.
6. **Write the deliverability note** โ€” explain in one line why this campaign protects sender reputation (e.g., "removes non-responders from future sends, preventing complaint-rate creep on a cold segment that would otherwise get mailed indiscriminately").
7. **Compliance tags on every email** โ€” even a bare plain-text question is a commercial email under CAN-SPAM if it's part of a marketing list; unsubscribe + postal address still required.
8. **Keep it genuinely plain-text-styled** โ€” no HTML template flourishes; the format's power depends on looking like a personal email, not a campaign (Halbert's A-pile principle, ยง4.1, applied to the modern inbox).

## Worked mini-example (19Keys / dormant CRM segment)

**Segment:** buyers with `last_engagement > 120 days` and no open/click in that window, from the 4,591-buyer CRM list.
**Email 1 (โ‰ค25 words):** Subject: "still building?" Body: "Hey [First] โ€” still working on leveling up your readiness? No worries either way, just closing out my list. โ€” 19Keys"
**Email 2 (day 5, non-responders only):** Subject: "one more try" Body: "Figured I'd check once more before I stop reaching out โ€” still interested in what we're building? Reply if so, no hard feelings if not."
**Sunset rule:** "Contacts with zero reply across both emails move to `suppressed_dead` segment in the CRM; no further marketing sends without a fresh explicit opt-in event."
**Deliverability note:** "Segments non-responders out of all future broadcast/flow sends, preventing complaint-rate creep on a 120+ day cold slice that would otherwise continue receiving daily/newsletter volume."

## How the verify gate applies

Per ยง8.1 `reactivation-campaign` rubric, checked for:
- **Target segment defined by an explicit inactivity rule** โ€” a vague segment description fails.
- **2-3 email arc**, starting with a nine-word-style question email โ€” **email 1 must be โ‰ค25 words total.**
- **Sunset rule stated** โ€” what happens to non-responders; missing this is a hard fail.
- **Deliverability note present** โ€” the one-line rationale for why this protects sender reputation.
- Compliance tags on all emails.
- Reviewer cross-check against ยง4.10: flag (even if the linter passes) any instance of this playbook being aimed at a hot/warm segment โ€” mechanically compliant but doctrinally wrong.